Causal Effects of Cannabis Legalization on Parents, Parenting, and Children: A Systematic Review

Abstract

Medical and recreational cannabis legalization lead to increased cannabis use among adults. There is concern that legalization has negative implications for minors via effects on parents. We conducted a systematic review of studies examining legalization in the United States. Web of Science, PsycInfo, and PubMed were searched through May, 2021, studies examining effects of legalization on maternal cannabis and other substance use during pregnancy and postpartum, perinatal outcomes, parental cannabis and other substance use and attitudes, parenting, and child outcomes were identified, and two independent reviewers extracted information on study designs, samples, and outcomes, and assessed classification of evidence and risk of bias. Forty-one studies met inclusion criteria; only 6 (15%) used the most causally informative study design (differences in differences). It is likely legalization increases maternal cannabis use during pregnancy and postpartum, parental cannabis use, and approval of adult cannabis use. Legalization may increase some adverse perinatal outcomes, though findings were inconsistent. It is likely legalization increases unintentional pediatric cannabis exposure. There is insufficient evidence for effects of legalization on child abuse and neglect, and there have been no studies examining effects of legalization on other aspects of parenting or on child adjustment. There is a critical lack of causally informative epidemiological studies examining effects of legalization on parenting and young children. Additional causally informative research is needed. Studies of parental cannabis use in a legal context are particularly needed. Commonsense guidelines must recognize the shifting national landscape around legalization while seeking to minimize potential harm to minors.

Introduction

Between 1996 and 2021, 36 states and the District of Columbia enacted medical cannabis legalization, and recreational cannabis legalization for adults age 21 years and older has now passed in 17 states and the District of Columbia¹. Proponents of legalization have cited several potential positive outcomes of cannabis legalization, including decreases in incarceration for drug-related crime², reduction in racial disparities in law enforcement³, possible reduction in harmful use of other substances^4–6, and generation of tax revenues³.

On the other hand, a widely voiced concern is potential for harm of legalization on minors. The American Academy of Pediatrics (AAP) opposes legalization and discourages cannabis use by adults in the presence of minors⁷. The American College of Obstetricians and Gynecologists (ACOG) discourages cannabis use during pregnancy and breastfeeding, and notes legalization may increase use by pregnant women⁸. These statements are informed by an extensive literature indicating that parents who use cannabis are more likely to have offspring who use cannabis^9–14 and theories suggesting similarity in parent-offspring cannabis use is due to socialization processes (e.g., parental modeling, problematic parenting)^15–18. Recent reviews indicate that, despite a substantial decrease in adolescents’ perceptions of risks of cannabis use since legalization³, legalization has not led to an increase in adolescent use¹⁹, though it has led to an increase in adult use^20–22. In contrast, there are fewer studies and no systematic reviews addressing the potential impact of legalization on younger children. The present study is a systematic review of the growing literature suggesting that legalization may lead to other outcomes that may affect minors negatively, including increased maternal cannabis or other substance use during pregnancy and postpartum, parental cannabis or other substance use, parenting impairment, and harmful effects on children.

Effects of Legalization on Adolescent and Adult Cannabis and Other Substance Use

Cannabis use is significantly higher in states with legalized medical and recreational cannabis²³. (We use the term “legalization” to refer to both medical and recreational legalization, noting the type of legalization when relevant.) States that have legalized cannabis also had higher rates of cannabis use prior to legalization, which, in conjunction with more permissive attitudes, may have led to legalization in the first place. Identifying causal influences of legalization requires causally informative study designs. In the “differences in differences” (DiD) approach²⁴, states with legalization are compared to states without legalization in difference in cannabis use before and after legalization. A recent systematic review and meta-analysis of DiD studies concludes that medical legalization does not increase adolescent cannabis use¹⁹. DiD studies of recreational legalization and adolescent use report similar findings. One study found recreational legalization increased cannabis use in younger, but not older, adolescents in Washington, but not Colorado²⁵. However, more recent studies found a decrease in adolescent cannabis use^26,27. Another study found no change in past month use/frequent use but an increase in cannabis use disorder²⁸.

In contrast, DiD studies find consistent evidence that medical legalization increases adult cannabis use and cannabis use disorders^20–22. Research using the DiD approach for recreational legalization and adult use is just beginning, with one recent study reporting increases in adult cannabis use and cannabis use disorder²⁸. These studies provide strong, converging evidence legalization may increase adult cannabis use and cannabis use disorder. Whether legalization has causal influences on other substance use among adults is unclear, with some studies suggesting decreases in alcohol use (e.g., decrease in traffic fatalities involving alcohol²⁹) and opiates (e.g., decrease in opioid pain reliever-related hospitalizations⁵) following legalization, but others suggesting increases in use of substances other than cannabissee ³⁰ for a review.

The Present Systematic Review: Effects of Legalization on Parental Cannabis and Other Substance Use, Parenting, and Young Children

In line with increasing adult cannabis use, nationwide, parents’ past-month use increased from 4.9% in 2002 to 6.8% in 2015³¹, suggesting that parents are not immune to potential effects of legalization and accompanying shifting cannabis use norms. However, there has not yet been a systematic review of studies examining effects of legalization on parental cannabis and other substance use, parenting, or young children. We conducted a systematic review addressing the hypothesis that medical and recreational legalization leads to outcomes that may negatively affect minors: increased maternal cannabis and other substance use during pregnancy and postpartum, and adverse prenatal, perinatal, and postnatal outcomes; increased parental cannabis and other substance use and more permissive attitudes toward cannabis and other substance use; parenting impairment, including increased child abuse or neglect; and other harmful effects on young children, including behavioral maladjustment and unintentional pediatric cannabis exposure. To determine the current evidence base for potentially causal effects of legalization on parental cannabis and other substance use, parenting, and young children, we also assessed classification of evidence (whether studies use the DiD approach; i.e., examination of pre-post legalization with state comparison) and risk of bias in the included studies.

Methods

Literature Search

Search strategy.

A systematic review of the literature examining effects of cannabis legalization on parental cannabis and other substance use, parenting, and young children was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines³² (see Table S1 for the PRISMA Checklist). Web of Science, PubMed, and PsycInfo databases were searched for studies indexed as published or in press through May, 2021. The literature search included journal articles, conference proceedings, and dissertations. Boolean search terms used were (((cannabis OR marijuana) AND legalization) OR “medical cannabis” OR “recreational cannabis” OR “medical marijuana” or “recreational marijuana”) AND (parent OR parents OR parenting OR prenatal OR perinatal OR neonatal OR postnatal OR pregnant OR pregnancy OR breastfeeding OR nursing OR lactation OR maltreatment OR neglect OR “child abuse” OR child OR children OR infant OR infants OR “pediatric exposure”) (see Figure S1 for the full electronic search strategy), which yielded 1,529 records. The literature search was not preregistered.

Inclusion and exclusion criteria.

Given the relatively sparse literature on the potential impact of legalization on younger children, the present study included a wide variety of outcomes that may affect minors negatively. Also, given the small number of studies in this literature, studies using a wide variety of designs (pre- and post-legalization with state comparison, pre- and post-legalization without state comparison, post-legalization with state comparison, and post-legalization without state comparison) were included. Duplicates were removed using Systematic Review Accelerator³³ and Rayyan³⁴, and 1,060 remaining study records were screened in Rayyan by one author (S.H.R.), who made initial determinations regarding exclusions of records if they (1) were reviews, commentaries, books, case studies, or animal studies, (2) did not consider a relevant outcome, (3) did not address legalization, or (4) were conference proceedings that were later published. Studies were included for full-text review if they (1) reported a primary research study on effects of medical or recreational legalization on maternal cannabis and other substance use during pregnancy and postpartum, prenatal, perinatal, and postnatal outcomes, parental cannabis and other substance use and attitudes toward cannabis and other substance use, parenting, or child outcomes, and (2) were conducted in the United States^†. Both authors independently conducted full-text reviews for 51 studies and reached decisions regarding eligibility for inclusion of 41 studies by consensus. The PRISMA flow diagram for the literature search, screening, and full-text review is in Figure 1.

Figure 1.

PRISMA flow diagram of the systematic literature search.

Study outcomes.

Prenatal, perinatal, and postnatal outcomes included maternal cannabis and other substance use during pregnancy or postpartum and prenatal, perinatal, and postnatal outcomes; parental and parenting outcomes included parental cannabis and other substance use, attitudes toward cannabis and other substance use, parenting, and child abuse or neglect; and child outcomes included child behavioral adjustment and children’s unintentional cannabis exposure.

Data extraction.

Both authors independently extracted data from 41 eligible studies: study citation, location, sample size, parent sex and age, child sex and age, parent/child race/ethnicity, study design, outcomes examined, and findings. Any disagreements in coding were resolved through discussion.

Classification of Evidence and Risk of Bias Assessment

Both authors independently assessed classification of evidence and risk of bias using a system adapted from the American Academy of Neurology criteria for rating causation studies³⁷ and the Risk of Bias in Non-randomized Studies - of Interventions (ROBINS-I) tool³⁸. Classification of evidence criteria yield classifications based on evidence for causation (Class I: samples assessed before and after legalization in states with and without legalization [i.e., the DiD approach], Class II: samples assessed before and after legalization in states with legalization, Class III: samples assessed after legalization in states with and without legalization, Class IV: samples assessed after legalization in states with legalization or samples assessed before and after legalization in states with legalization unspecified). The ROBINS-I tool yields ratings (low: comparable to well-performed randomized trial with regard to domain, moderate: sound for a non-randomized study with regard to domain but cannot be considered comparable to a well-performed randomized trial, serious: has some important problems, critical: too problematic to provide any useful evidence on the effects of intervention) on biases due to confounding, selection of participants, classification of the intervention, deviations from the intervention, missing data, measurement of outcomes, and analysis and reporting of results, allowing an overall judgement of risk of bias for each study. Any rating discrepancies were resolved by consensus. Classification of the evidence and risk of bias assessment rating forms used for this systematic review are in Tables S2, S3, S4.

Results

This systematic review summarizes results of the 41 studies that met inclusion/exclusion criteria (see Table 1; one study examined both maternal cannabis and other substance use during pregnancy and parental cannabis and other substance use³⁹). The 20 studies examining prenatal, perinatal, and postnatal outcomes included maternal cannabis and other substance use and disorder during pregnancy and postpartum^39–54, cannabis and other substance treatment admissions during pregnancy⁵⁵, maternal attitudes toward cannabis use during pregnancy and postpartum⁴¹, and adverse maternal and child prenatal, perinatal, and postnatal outcomes^{43,46,52,56–58}. The 12 studies examining parental and parenting outcomes included parental cannabis and other substance use and disorder^39,59–62, parental attitudes toward cannabis and other substance use^{45,59,61,63–65}, knowledge of cannabis law^61,64, cannabis storage practices in the home⁶⁶, and child abuse and neglect^67,68. The 10 studies examining child outcomes all included unintentional pediatric cannabis exposure^69–78; no other child outcomes, including child behavioral adjustment, were examined. The studies were journal articles with the exception of 2 conference presentations^40,48 and 2 dissertations^45,69.

Table 1.

Overview of Studies Included in the Systematic Review

Study	Location	Develop-mental period and/or age of children	Sample description/size	Sample race/ethnicity	Study design	Outcome	Overview	Evidence class
Prenatal, perinatal, postnatal outcomes
Ali et al. (in press)39‡	States with medical cannabis legalization, states without medical cannabis legalization	Prenatal	11,700 pregnant women	14.0% non-Hispanic Black 76.7% non-Hispanic white 6.4% non-Hispanic other 19.2% Hispanic	Pre-post medical cannabis legalization with state comparison	Cannabis and opioid use past month and past year, cannabis and opioid use disorder past year, frequency of cannabis and opioid use, initiation of cannabis and opioid use during pregnancy assessed via self-report	Significant increase pre-post medical cannabis legalization in past-year cannabis use frequency during pregnancy assessed via self-report in states with medical cannabis legalization compared to states without; non-significant decrease pre-post medical cannabis legalization in past-month cannabis use, past-year cannabis use, past-year cannabis use disorder, past-year cannabis initiation during pregnancy assessed via self-report in states with medical cannabis legalization compared to states without, but significant decrease in past-year cannabis initiation during pregnancy in states granting legal protection for medical cannabis dispensaries; significant increase pre-post medical cannabis legalization in past-year opioid misuse frequency during pregnancy assessed via self-report in states with medical cannabis legalization compared to states without; non-significant increase pre-post medical cannabis legalization in past-month opioid misuse, non-significant decrease in past-year opioid misuse, past-year opioid use disorder, past-year opioid initiation during pregnancy assessed via self-report in states with medical cannabis legalization compared to states without	I
Allshouse et al. (2016)40	Colorado	Prenatal	743 pregnant women (371 before and 372 after legalization)	Not reported	Pre-post recreational cannabis legalization without state comparison	Cannabis use during pregnancy assessed via urine toxicology and self-report	Non-significant increase pre-post recreational cannabis legalization in cannabis use during pregnancy assessed via urine toxicology and self-report	II
Barbosa-Leiker et al. (2020)41	Washington state	Prenatal, postpartum (through infant age 3 months)	19 pregnant and postpartum women	Not reported (“predominant-ly Caucasian”)	Post-recreational cannabis legalization without state comparison	Cannabis use during pregnancy and postpartum, attitudes toward cannabis use assessed via self-report	Continued daily cannabis use during prenatal and postpartum periods assessed via self-report in focus group by women; women reported changing use due to legal considerations, using for health management, continuing to evaluate their use, often receiving mixed messages from healthcare providers, frustration over lack of information on health risks	IV
Crume et al. (2018)42	Colorado	Prenatal, perinatal, postpartum (through infant age 4 months)	3,207 pregnant women	Not reported	Post-recreational cannabis legalization without state comparison	Cannabis use during pregnancy and postpartum assessed via self-report	5.7% cannabis use during pregnancy, 5.0% cannabis use during breastfeeding assessed via self-report	IV
Gnofam et al. (2020)43	Colorado	Prenatal, perinatal	2,392 pregnant women (1,165 before and 1,227 after legalization)	Before legalization 16.1% Black 0.4% Native American/Alaskan Native 3.2% Asian 0.6% Native Hawaiian/other Pacific Islander 48.2% white 0.1% more than one race 27.6% other 32.5% Hispanic After legalization 16.7% Black 0.4% Native American/Alaskan Native 4.2% Asian 0.2% Native Hawaiian/other Pacific Islander 48.8% white 0.2% more than one race 23.4% other 32.5% Hispanic	Pre-post opening of recreational cannabis dispensaries without state comparison	Cannabis and other substance use during pregnancy assessed via urine toxicology, meconium testing, and self-report; preeclampsia, infant growth restriction, preterm birth, stillbirth, infant or maternal death	Significant increase pre-post opening of recreational cannabis dispensaries in cannabis use during pregnancy assessed via urine toxicology and self-report; non-significant increase pre-post opening of recreational cannabis dispensaries in cannabis use during pregnancy assessed via meconium testing; non-significant increase pre-post opening of recreational cannabis dispensaries in tobacco, alcohol, opioid, other drug use during pregnancy assessed via urine toxicology, meconium testing and self-report; significantly greater incidence of infant growth restriction, no significant differences for other perinatal outcomes	II
Grant et al. (2018)44	Washington state	Prenatal, postpartum (through child age 3 years)	1,359 pregnant women with substance use disorders (997 before and 362 after legalization)	Before legalization 12.1% Black 17.9% Native American 58.8% white 2.1% other 9.1% Hispanic After legalization 9.7% Black 20.4% Native American 58.6% white 8.6% other 9.9% Hispanic	Pre-post recreational cannabis legalization without state comparison	Cannabis and other substance use during pregnancy and postpartum assessed via self-report	Non-significant increase pre-post recreational cannabis legalization in cannabis use at program enrollment (during pregnancy or postpartum periods) assessed via self-report; significant increase pre-post recreational cannabis legalization in heroin, methadone, and other opioid use at program enrollment (during pregnancy or postpartum periods) assessed via self-report; significant decrease pre-post recreational cannabis legalization in cocaine, barbiturate, and alcohol use at program enrollment (during pregnancy or postpartum periods) assessed via self-report; significant increase pre-post recreational cannabis legalization in cannabis, alcohol, methadone, and other opioid use at program exit (during postpartum period) assessed via self-report	II
Jones et al. (2017)45	22 states with medical cannabis legalization, 28 states without medical cannabis legalization	Prenatal	758 pregnant women (306 from states with legalization and 452 from states without legalization)	14.2% non-Hispanic Black 2.6% Native American/ Alaskan Native 1.3% Native Hawaiian/other Pacific Islander 4.5% Asian 54.6% non-Hispanic white 3.3% more than one race 19.4% Hispanic	Post-medical cannabis legalization with state comparison	Cannabis use past month and cannabis use past year assessed via self-report	Non-significant difference in past month cannabis use or past year cannabis use during pregnancy assessed via self-report by pregnant women in states with medical cannabis legalization compared to states without	III
Lee et al. (2020)46	California	Prenatal, perinatal	466 pregnant women (279 before and 187 after legalization)	11.2% Black 17.9% Native American/ Alaskan Native 20.4% white 18.2% other 49.1% Hispanic	Pre-post recreational cannabis legalization without state comparison	Cannabis use during pregnancy assessed via urine toxicology, other substance use during pregnancy assessed via clinic records; small for gestational age, preterm birth, NICU admission, Apgar score	Significant increase pre-post recreational cannabis legalization in cannabis use during pregnancy assessed via urine toxicology; no significant differences for other substance use assessed via clinic records; no significant difference for perinatal outcomes	II
Lockwood et al. (2019)56	Colorado	Perinatal	269,922 mother-infant dyads (125,599 before and 144,323 after legalization)	5.0% Black 82.0% white 11.0% other	Pre-post recreational cannabis legalization without state comparison	Small for gestational age, preterm birth, NICU admission, low birth weight, infant death	Significant increase pre-post recreational cannabis legalization in NICU admissions; no significant differences for other perinatal outcomes	II
Meinhofer et al. (2019)55	21 states with medical cannabis legalization, 27 states without medical cannabis legalization	Prenatal	282,955 pregnant women and 6,273,520 non-pregnant women	Not reported	Pre-post medical cannabis legalization with state comparison	Cannabis treatment admissions, other substance treatment admissions	Significant increase pre-post medical cannabis legalization in cannabis treatment admissions by pregnant women in states with medical cannabis legalization compared to states without, with largest difference in states granting legal protection for medical cannabis dispensaries; no significant difference in cannabis treatment admissions by non-pregnant women in states with medical cannabis legalization compared to states without; significant increase in alcohol and cocaine treatment admissions by pregnant women in states with medical cannabis legalization compared to states without; no significant difference in opioid treatment admissions by pregnant women in states with medical cannabis legalization compared to states without	I
Metz et al. (2019)47	Colorado	Prenatal	116 pregnant women	12.1% Black 3.4% Asian 71.6% white 12.1% other	Post-recreational cannabis legalization without state comparison	Cannabis use during pregnancy assessed via umbilical cord testing, cannabis use past month, report of cannabis use to health care providers assessed via self-report	22% cannabis use assessed via umbilical cord testing, 6% past month cannabis use during pregnancy assessed via self-report, 2.6% reported cannabis use to health care providers assessed via self-report	IV
Petrova et al. (2021)57	States with medical cannabis legalization, states without medical cannabis legalization	Perinatal	84,518,136 infants (27,723,116 from states with legalization, 56,795,020 from states without legalization)	15.2% Black 78.8% white 6.0% other 20.9% Hispanic	Pre-post medical cannabis legalization with state comparison	Preterm birth, birth weight, Apgar score	Significant increase pre-post medical cannabis legalization in birth weight in states with medical cannabis legalization compared to states without; no significant differences for other perinatal outcomes	I
Rodriguez et al. (2016)48	Colorado	Prenatal	1,698 pregnant women from adolescent-specific prenatal care program (1,148 before and 550 after legalization)	Not reported	Pre-post recreational cannabis legalization without state comparison	Cannabis use during pregnancy assessed via urine toxicology	Significant increase pre-post recreational cannabis legalization in cannabis use during pregnancy assessed via urine toxicology	II
Short et al. (2020)49	3 states with medical or recreational cannabis legalization (Maine, New Mexico, Vermont), 3 states without medical or recreational cannabis legalization (Louisiana, Wisconsin, Wyoming)	Preconception	5,676 women	18.0% Black 71.6% white 10.4% other	Post-medical or recreational cannabis legalization with state comparison	Cannabis use past month preconception assessed via self-report	Significantly greater cannabis use during preconception period assessed via self-report in states with medical or recreational cannabis legalization compared to states without	III
Siega-Riz et al. (2020)58	Colorado and Washington state	Perinatal	1,347,916 infants (576,369 from Colorado and 771,547 from Washington)	Colorado 4.3% Black 59.8% white 6.8% other 29.1% Hispanic Washington 5.0% Black 63.6% white 12.6% other 18.8% Hispanic	Pre-post recreational cannabis legalization without state comparison	Small for gestational age, preterm birth, congenital anomalies	Significant increase pre-post recreational cannabis legalization in prevalence of congenital anomalies in Colorado and Washington; significant increase in preterm births in Colorado but not Washington; non-significant increase in small for gestational age in Colorado and Washington	II
Skelton et al. (2020)50	3 states with recreational cannabis legalization (Alaska, Colorado, Washington), 3 states without recreational cannabis legalization (Maine, Michigan, New Hampshire)	Preconception, prenatal, postpartum (through infant age 9 months)	7,258 women (4,076 from states with legalization and 3,182 from states without legalization)	9.8% Black 1.6% Native American/ Alaskan Native 6.1% Asian 74.1% white 8.5% other 14.9% Hispanic	Post-recreational cannabis legalization with state comparison	Cannabis use past year preconception, during pregnancy, postpartum assessed via self-report	Significantly greater cannabis use during preconception, prenatal, and postpartum periods assessed via self-report in states with recreational cannabis legalization compared to states without	III
Skelton et al. (2021)51	2 states with recreational cannabis legalization (Alaska, Maine), 2 states without recreational cannabis legalization (New Hampshire, Vermont)	Preconception, prenatal, postpartum (through infant age 6 months)	23,082 women preconception (17,616 from states with legalization and 5,466 from states without legalization), 23,859 pregnant women (18,217 from states with legalization and 5,642 from states without legalization), 26,610 women postpartum (19,780 from states with legalization and 6,830 from states without legalization)	Preconception 70.4% non-Hispanic white Pregnant 71.0% non-Hispanic white Postpartum 72.1% non-Hispanic white	Pre-post recreational cannabis legalization with state comparison	Cannabis use past year preconception, during pregnancy, postpartum assessed via self-report	Significant increase pre-post recreational cannabis legalization in cannabis use during preconception and postpartum periods assessed via self-report in states with recreational cannabis legalization compared to states without; non-significant increase pre-post recreational cannabis legalization in cannabis use during pregnancy assessed via self-report in states with recreational cannabis legalization compared to states without	I
Straub et al. (in press)52	Washington state	Perinatal	5,343 pregnant women (1,610 before and 3,733 after legalization)	14.5% Black 57.4% white	Pre-post recreational cannabis legalization without state comparison	Cannabis and illicit drug use during pregnancy assessed via urine toxicology, alcohol and cigarette use during pregnancy assessed via self-report; small for gestational age, low birth weight	Non-significant difference pre-post recreational cannabis legalization in cannabis or illicit drug use during pregnancy assessed via urine toxicology; significant increase pre-post recreational cannabis legalization in alcohol use during pregnancy assessed via self-report; significant decrease pre-post recreational cannabis legalization in cigarette use during pregnancy assessed via self-report; non-significant increase pre-post recreational cannabis legalization in small for gestational age; significant increase pre-post recreational cannabis legalization in low birth weight	II
Volkow et al. (2019)53	United States	Prenatal	467,100 pregnant women	Not reported	Pre-post recreational cannabis legalization without state comparison	Cannabis use past month assessed via self-report	Significant increase pre-post recreational cannabis legalization in past month cannabis use during pregnancy assessed via self-report	IV
Young-Wolff et al. (2021)54	California	Prenatal	35,195 pregnant women	6.2% Black 26.9% Asian/Pacific Islander 36.1% Non-Hispanic white 4.3% more than one race, other, unknown 26.5% Hispanic	Post-recreational cannabis legalization without state comparison	Cannabis use during pregnancy assessed via urine toxicology and self-report	8.1% cannabis use during pregnancy assessed via urine toxicology or self-report; significantly higher odds of cannabis use for living within a 15-minute drive of cannabis retailers, significantly lower odds of cannabis use for longer drive-time from cannabis retailers	IV
Parental and parenting outcomes
Ali et al. (in press)39‡	States with medical cannabis legalization, states without medical cannabis legalization	Infancy (< 3 years), childhood, adolescence (< 18 years)	47,700 mothers of infants; 117,600 mothers of children/adolescents	Parents of infants 3.7% non-Hispanic Black 77.6% non-Hispanic white 5.7% non-Hispanic other 20.7% Hispanic Parents of children 14.2% non-Hispanic Black 77.6% non-Hispanic white 5.7% non-Hispanic other 18.3% Hispanic	Pre-post medical cannabis legalization with state comparison	Parental cannabis and opioid use past month and past year, cannabis and opioid use disorder past year, frequency of cannabis and opioid use, initiation of cannabis and opioid use assessed via self-report	Parents of infants Significant increase pre-post medical cannabis legalization in parental past-year cannabis use frequency assessed via self-report in states with medical cannabis legalization compared to states without; non-significant decrease pre-post medical cannabis legalization in parental past-month cannabis use, past-year cannabis use, past-year cannabis initiation assessed via self-report in states with medical cannabis legalization compared to states without; non-significant increase in parental past-year cannabis use disorder in states with medical cannabis legalization compared to states without; significant decrease pre-post medical cannabis legalization in parental past-year opioid misuse frequency assessed via self-report in states with medical cannabis legalization compared to states without, with larger decrease in states granting legal protection for medical cannabis dispensaries; non-significant increase pre-post medical cannabis legalization in parental past-month opioid misuse, past-year opioid misuse, past-year opioid initiation, but significant increase in parental past-year opioid initiation in states granting legal protection for medical cannabis dispensaries; non-significant decrease in past-year opioid use disorder assessed via self-report in states with medical cannabis legalization compared to states without Parents of children Significant increase pre-post medical cannabis legalization in parental past-month cannabis use, past-year cannabis use frequency assessed via self-report in states with medical cannabis legalization compared to states without, with larger decrease in states granting legal protection for medical cannabis dispensaries; non-significant increase pre-post medical cannabis legalization in parental past-year cannabis use, past-year cannabis use disorder assessed via self-report in states with medical cannabis legalization compared to states without; non-significant decrease pre-post medical cannabis legalization in parental past-year cannabis initiation assessed via self-report in states with medical cannabis legalization compared to states without; significant increase pre-post medical cannabis legalization in parental past-year opioid misuse assessed via self-report in states with medical cannabis legalization compared to states without; significant decrease pre-post medical cannabis legalization in parental past-year opioid misuse frequency assessed via self-report in states with medical cannabis legalization compared to states without, with larger decrease in states granting legal protection for medical cannabis dispensaries; non-significant increase pre-post medical cannabis legalization in parental past-month opioid misuse, past-year opioid initiation assessed via self-report in states with medical cannabis legalization compared to states without; non-significant decrease in past-year opioid use disorder assessed via self-report in states with medical cannabis legalization compared to states without	I
Brooks-Russell et al. (2020)66	Colorado	Infancy, childhood, adolescence (range 1 to 14 years)	3,775 caregivers	Not reported	Post-recreational cannabis legalization without state comparison	Cannabis storage practices in the home assessed via self-report	Significant increase post recreational cannabis legalization in cannabis in homes with children assessed via self-report; most stored cannabis in locations inaccessible to children, majority stored cannabis in a locked container	IV
Eisenberg et al. (2019)63	Washington state	Childhood, adolescence (range 8 to 15 years)	54 parents	37.0% Black 7.0% Native American/Alaskan Native 17.0% Asian 39.0% white	Post-recreational cannabis legalization without state comparison	Attitudes toward adolescent cannabis use assessed via self-report	Attitudes toward adolescent cannabis use assessed via self-report in focus group by parents of adolescents; parents reported trouble reconciling societal and personal norms, deciding whether to disclose personal use to adolescence; belief that underage use is unacceptable but adolescent experimentation is likely; talking to their adolescents about cannabis; setting household guidelines but challenges monitoring adolescents, imposing consequences for cannabis use; desire to learn strategies, including factual information about cannabis and parenting skills	IV
Epstein et al. (2020)59	5 states with recreational cannabis legalization (Alaska, California, Colorado, Oregon, Washington), 24 states without recreational cannabis legalization	Not given	426 parents, 242 caregivers	21.0% Black 6.0% Native American/Alaskan native 20.0% Asian 52.0% white	Pre-post recreational cannabis legalization with state comparison	Parental cannabis use frequency past year, attitudes toward cannabis use assessed via self-report	Significant increase pre-post recreational cannabis legalization in parental past-year cannabis use frequency assessed via self-report in states with recreational cannabis legalization compared to states without, with increases for new and existing cannabis users but larger increases for new users; significant increase pre-post medical cannabis legalization in approval of adult cannabis use assessed via self-report in states with recreational cannabis legalization compared to states without	I
Freisthler et al. (2015)67	California	Childhood (<= 12 years)	3,023 parents/legal guardians and 2,909 children	4.8% non-Hispanic Black 49.4% non-Hispanic white 15.2% more than one race, other 30.6% Hispanic	Post-medical cannabis legalization without state comparison	Physical abuse, supervisory neglect, physical neglect assessed via self-report	Significant positive association between density of medical cannabis dispensaries and frequency of physical abuse assessed via self-report; non-significant association with frequency of supervisory or physical neglect assessed via self-report	IV
Freisthler et al. (2020)68	California	Childhood	2,342 census tracts (average of 991 children each tract)	8.1% Black 47.4 Hispanic	Post-medical cannabis legalization without state comparison	Abuse or neglect referrals for child protective services investigations	Significant positive association between density of medical cannabis dispensaries and referrals for child protective services investigations; non-significant association after adjusting for alcohol outlet density and social disorganization	IV
Goodwin et al. (2021)60	United States	Childhood, adolescence (< 18 years)	287,624 adults with children living in the home (22,308 in 2017)	Not reported	Pre-post medical and recreational cannabis legalization with state comparison; post-medical and recreational cannabis legalization with state comparison (in 2017)	Adult cannabis use past month and daily cannabis use assessed via self-report	Significant increase pre-post medical and recreational cannabis legalization in adult past-month cannabis use and daily cannabis use assessed via self-report in states with medical or recreational cannabis legalization compared to states without; greater past-month cannabis use and daily cannabis use in states with medical and recreational cannabis legalization compared to states without (in 2017)	I
Jones et al. (2020)79	Washington state	Childhood, adolescence (range 8 to 15 years)	54 parents	37.0% Black 1.9% Native American 16.7% Asian 38.9% white	Post-recreational cannabis legalization without state comparison	Attitudes toward adolescent cannabis use assessed via self-report	Attitudes toward adolescent cannabis use assessed via self-report in focus group by parents of adolescents; parents reported not wanting their adolescents to use cannabis; concern of increased adolescent exposure to cannabis, increased risk of adolescent cannabis use after cannabis legalization; greater need to monitor adolescents; particular concern about edible cannabis products	IV
Kosterman et al. (2016)61	Washington state	Childhood, adolescence (<= 19 years)	395 parents	27.0% Black 5.0% Native American/Alaskan Native 22.0% Asian 45.0% white	Pre-post recreational legalization without state comparison	Parental cannabis use past year, attitudes toward cannabis use, attitudes toward adolescent cannabis use, knowledge of cannabis law assessed via self-report	Significant increase post recreational cannabis legalization in parental past year cannabis use frequency, cannabis use disorder assessed via self-report; significant increase pre-post recreational cannabis legalization in approval of adult cannabis use assessed via self-report; significant decrease pre-post recreational cannabis legalization in perceived harm of adult cannabis use assessed via self-report; parents reported disapproval of cannabis use while parenting, adolescent cannabis use; parents uncertain about age of legal cannabis use	II
Mason et al. (2015)64	Washington state	Adolescence (16 years)	115 parents	21.0% Black 51.0% white	Post-recreational cannabis legalization without state comparison	Attitudes toward adolescent cannabis use, knowledge of cannabis law assessed via self-report	Attitudes toward adolescent cannabis use assessed via self-report; parents reported little change in approval of adult or adolescent cannabis use or likelihood of cannabis use; parents with lifetime cannabis use reported greater approval of adult cannabis use and likelihood of cannabis use compared to parents without lifetime cannabis use; parents reported talking to their adolescents about cannabis legalization but infrequently; parents uncertain about age of legal cannabis use	IV
Rusby et al. (2018)62	Oregon	Adolescence (14 years)	343 parents and 444 adolescents	39% Hispanic 61% non-Hispanic (“predominant-ly white”)	Pre-post recreational cannabis legalization without state comparison	Parental cannabis use past year assessed via self-report	Non-significant increase pre-post recreational cannabis legalization in parental cannabis use assessed via self-report	II
Thurstone et al. (2013)65	Colorado	Not given	11 parents	18.8% Black 3.1% Native American/Alaskan Native 59.4% white 18.8% other 12.5% Hispanic	Post-medical cannabis legalization without state comparison	Attitudes toward cannabis use, attitudes toward adolescent cannabis use assessed via self-report	Attitudes toward cannabis use assessed via self-report in focus group by parents of adolescents; parents reported cannabis use helped them relax and stay calm when parenting; trouble deciding whether to disclose personal use to adolescents; disapproval of adolescent cannabis use, belief cannabis use is preferable to other substance use	IV
Child outcomes
Close et al. (2019)69	Washington state	Infancy, childhood (< 12 years of age)	50 children	Not reported	Post-recreational cannabis legalization without state comparison	Cannabis exposures reported to regional poison centers	Most cannabis exposures involved ingestion, occurred at home, and caused no or minor health effects; source of cannabis was commonly parent or grandparent	IV
Dean et al. (2021)70	Michigan	Infancy, childhood, adolescence (< 18 years of age)	426 children	Not reported	Post-medical and pre-post recreational cannabis legalization without state comparison	Cannabis exposures reported to regional poison centers	Increase in post-medical and pre-post recreational cannabis legalization in cases of unintentional cannabis exposure reported to poison center	II†
Onders et al. (2016)71	United States	Infancy, childhood (< 6 years of age)	1969 children	Not reported	Post-medical cannabis legalization with state comparison	Cannabis exposures reported to regional poison centers	Significant increase post-medical cannabis legalization in cases of unintentional cannabis exposure reported to poison centers in states with medical cannabis legalization compared to states without	III
Thomas et al. (2018)72	Washington state	Infancy, childhood (median age = 21 months; range 17 months to 7 years)	17 children (8 before and 9 after legalization)	Not reported	Pre-post recreational cannabis legalization without state comparison	Cannabis exposure assessed via urine toxicology	Significant increase pre-post recreational cannabis legalization in cases of unintentional cannabis exposure assessed via urine toxicology	II
Thomas et al. (2019)73	Washington state	Infancy, childhood (median age = 2 years; range 0 to 9 years)	161 children (31 before and 130 after legalization)	Not reported	Pre-post recreational cannabis legalization, pre-post opening of recreational cannabis dispensaries without state comparison	Cannabis exposures reported to regional poison center	Significant increase pre-post recreational cannabis legalization in cases of unintentional cannabis exposure reported to poison center; significant increase pre-post opening of recreational cannabis dispensaries in cases of unintentional cannabis exposure reported to poison center	II
Wang et al. (2013)74	Colorado	Infancy, childhood (median age = 2 years)	1,378 children (790 before and 588 after legalization)	Not reported	Pre-post medical cannabis decriminalization without state comparison	Cannabis exposures at children's hospital visits	Significant increase pre-post recreational cannabis legalization in cases of unintentional cannabis exposure at children’s hospital visits	II
Wang et al. (2014)75	9 states with decriminalized medical cannabis (i.e., laws passed before 2005), 5 states transitioning to medical cannabis legalization (i.e., legislation between 2005 and 2011), 36 states without medical cannabis legalization	Infancy, childhood (median = 2 years)	985 children (396 from decriminalized states, 93 from transitional states, 496 from nonlegal states)	Not reported	Post cannabis decriminaliza-tion with state comparison	Cannabis exposures reported to national poison system	Significantly more cases of unintentional cannabis exposure reported to poison system in decriminalized states compared to nonlegal states; non-significant difference in cases of unintentional cannabis exposure reported to poison system in transitional states	III
Wang et al. (2016)76	Colorado	Infancy, childhood (median age = 2 years)	244 children (81 from a children's hospital and 163 from a poison center)	Not reported	Pre-post recreational cannabis legalization without state comparison	Cannabis exposures at children's hospital visits and reported to regional poison center	Significant increase pre-post recreational cannabis legalization in cases of unintentional cannabis exposure at children’s hospital visits and reported to poison center	II
Wang et al. (2019)77	Colorado	Infancy, childhood (median age = 3 years)	168 children (52 from a children's hospital and 116 from a poison center)	Not reported	Post-recreational cannabis legalization without state comparison	Cannabis exposures at children's hospital visits and reported to regional poison center	Significant increase post recreational cannabis legalization in cases of unintentional cannabis exposure at children’s hospital visits and reported to poison center	IV
Whitehill et al. (2019)78	Massachusetts	Infancy to adolescence (0 to 4 years: 10%; 5 to 9 years: 2%; 10 to 14 years: 6%; 15 to 19 years: 82%)	218 children	Not reported	Pre-post medical cannabis legalization without state comparison	Cannabis exposures reported to regional poison center	Significant increase pre-post medical cannabis legalization in cases of unintentional cannabis exposure reported to poison center	II

Note. Studies included in the systematic review, sample characteristics, study design, outcome of interest, overview of key results, and evidence class. Classification of the evidence ratings adapted from the American Academy of Neurology classification of evidence schemes criteria for rating causation studies³⁷. Class I = cohort survey with prospective data collection (pre-post cannabis legalization in state[s] with legalization with comparison with state[s] without legalization, i.e., DiD approach). Class II = cohort study with retrospective data collection or case-control study (pre-post cannabis legalization in state[s] with legalization without comparison with state[s] without legalization). Class III = cohort or case-control study designs (post-cannabis legalization in state[s] with legalization with comparison with state[s] without legalization). Class IV = studies not meeting Class I, II, or III criteria.

^‡Non-mutually exclusive samples of pregnant women, mothers of infants, mothers of children/adolescents.

Study Characteristics

Of the 20 studies examining effects of legalization on prenatal, perinatal, and postnatal outcomes, 4 (20%) studies used pre-post legalization with state comparison (i.e., DiD)^39,51,55,57. Eight (40%) studies used pre-post legalization without state comparison^{40,43,44,46,48,52,53,56,58} (one study compared across states with and without legalization⁵³), 3 (15%) studies used post-legalization with state comparison^45,49,50, and 5 (25%) studies used post-legalization without state comparison^41,42,47,54. The majority (n = 15, 75%) examined recreational legalization^{40–44,46–48,50–54,56,58}.

Of the 12 studies examining effects on parental and parenting outcomes, 3 (25%) studies used pre-post legalization with state comparison^39,59,60. Two (17%) studies used pre-post legalization without state comparison^61,62, and no (0%) studies used post-legalization with state comparison. The majority (n = 7, 58%) used post-legalization without state comparison^63–68,79. The majority (n = 8, 67%) examined recreational legalization^{59–64,66,79}.

Of the 10 studies examining effects on child outcomes, no (0%) studies used pre-post legalization with state comparison. Six (60%) studies used pre-post legalization without state comparison^{70,72–74,76,78‡}, 2 (20%) studies used post-legalization with state comparison^71,75, and 2 (20%) studies used post-legalization without state comparison^69,77. The majority (n = 6, 60%) examined recreational legalization^{69,70,72,73,76,77}.

Prenatal, Perinatal, and Postnatal Outcomes

Sixteen studies examined maternal cannabis use and problematic use (i.e., cannabis use disorder, treatment admissions) during pregnancy and postpartum, assessed using biodetection and self-report^{39–48,50–55} (5 [31%] studies conducted by two groups^{40,43,47,50,51}). There was evidence of post-legalization increases in cannabis use, cannabis use disorder, and cannabis treatment admissions during the preconception period, pregnancy, and breastfeeding in the 9 studies using the strongest designs (i.e., pre-post legalization with state comparison [i.e., DiD], pre-post legalization without state comparison)^{39,40,43,44,46,48,51,52,55}, which was significant in 6 studies^{39,43,46,48,51,55}. One possible explanation for this increase is that parents are simply more likely to report use post-legalization given reduced stigma around use and concerns regarding legal ramifications. However, two pre-post legalization studies without state comparison used both biodetection and self-report and reported increases using both methods^40,43. Moreover, one post-legalization study found higher rates of use with umbilical cord testing than self-report (22% versus 6%), suggesting stigma and concerns regarding legal ramifications remain for parents post-legalization⁴⁷. Based on classification of evidence for these 16 studies, it is likely that legalization increases cannabis use and problematic use during the preconception, pregnancy, and postpartum periods (3 Class I studies^39,51,55, 6 Class II studies^{40,43,44,46,48}). However, risk of bias assessment judged even the studies using the strongest designs at serious to critical risk of bias due to confounding, selection, and classification (see Figure 2).

Figure 2.

Rating of risk of bias in studies included in the systematic review. Risk of bias ratings adapted from Risk Of Bias In Non-Randomized Studies - of Interventions (ROBINS-1) tool for assessing risk of bias from studies that did not use randomization³⁸.

Six studies examined maternal other substance use and problematic use (i.e., other substance use disorder, treatment admissions) during pregnancy and postpartum^{39,43,44,46,52,55}. Results were mixed, with some studies finding increases in use of other substances after legalization^39,44,52,55, others finding decreases^44,52, and some finding no differences^39,43,46,55, with inconsistent findings across substances (e.g., alcohol, opioids). Based on classification of evidence for these 6 studies, there is insufficient evidence regarding effects of legalization on other substance use during pregnancy and postpartum (2 Class I studies^39,55, 4 Class II studies^43,44,46,52). Moreover, risk of bias assessment judged these studies at serious risk of bias due to confounding, selection, and classification, with 1 exception at moderate risk of bias⁵⁵ (see Figure 2).

Two studies examined maternal attitudes toward or reasons for cannabis use during pregnancy and postpartum, assessed using self-report^41,47. After legalization, women reported changing cannabis use due to legal considerations, using cannabis for health management (e.g., morning sickness, pain, sleep), mixed messages from healthcare providers, and frustration over lack of information on health risks⁴¹. Only a minority reported use to health care providers⁴⁷. Based on classification of evidence for these 2 studies, there is insufficient evidence regarding effects of legalization on attitudes toward cannabis use during pregnancy and postpartum (2 Class IV studies^41,47). Moreover, risk of bias assessment judged these studies at critical risk of bias due to confounding, selection, and reporting results (see Figure 2).

Six studies examined perinatal and postnatal outcomes^{43,46,52,56–58}. There was some evidence of increase in adverse perinatal and postnatal outcomes, which was significant in 4 studies for infant growth restriction⁴³, preterm birth⁵⁸, NICU admissions⁵⁶, low birth weight⁵², and congenital abnormalities⁵¹, but non-significant in 6 studies for small for gestational age, preterm birth, NICU admissions, Apgar score, low birth weight, congenital anomalies, stillbirth, and infant or maternal death^{43,46,52,56–58}; 1 study found a small but significant increase in birth weight⁵⁷. Based on classification of evidence for these 6 studies, it is possible legalization increases some adverse perinatal and postnatal outcomes, though findings were inconsistent (1 Class I study⁵⁷, 5 Class II studies^{43,46,52,56,58}). Moreover, risk of bias assessment judged these studies at serious risk of bias due to confounding, selection, and classification (see Figure 2).

Parental and Parenting Outcomes

Twelve studies examined parental and parenting outcomes^{39,59–65,66,67,68,79} (five [46%] studies conducted by one group^{59,61,63,64,79}). Five studies examined parental cannabis use (or use by adults with children living in the home), assessed using self-report^{39,59,60–62}. There was evidence of an increase in parental use, which was significant in 4 studies for cannabis use^39,59–61 and cannabis use disorder⁶¹. One study examined parental other substance use, assessed using self-report³⁹. There was evidence of both increases and decreases in opioid initiation and misuse, with a significant decrease in opioid misuse among mothers of infants and children, but a significant increase in opioid initiation among mothers of infants³⁹. Six studies examined parental attitudes toward cannabis and adult and adolescent use, assessed using self-report^{59,61,63–65,79}. There was evidence of a significant increase in parental approval of adult use^59,61 and a significant decrease in perceived harm of adult use⁶¹. After legalization, parents reported disapproval of use while parenting⁶¹ but also using cannabis to help them relax and stay calm when parenting⁶⁵. Parents reported concern about increased adolescent exposure to cannabis and increased adolescent use, disapproval of adolescent use but also likelihood of experimentation, talking with adolescents about cannabis, and challenges monitoring adolescent use^{61,63–65,79}. Two studies examined cannabis storage in the home^65,66, with most parents reporting safe storage practices. A sizeable minority of parents was uncertain about the age of legality (i.e., believed it was 18 years when it was 21 years^61,64). No other aspects of parenting or effects of legalization on parenting were examined. Based on classification of evidence for these 10 studies, it is likely legalization increases parental cannabis use and parental approval of adult use (3 Class I studies^39,59,60, 2 Class II studies^61,62) but there is insufficient evidence regarding effects of legalization on other substance use (1 Class I study³⁹), parental approval of adolescent cannabis or other substance use, cannabis storage practices in the home, or parenting (5 Class IV studies^63–66,79). Moreover, risk of bias assessment judged even the studies using the strongest designs at serious to critical risk of bias due to confounding, selection, measurement, and reporting results, with 2 exceptions at moderate risk of bias^59,60 (see Figure 2).

Two studies examined child abuse and neglect, assessed using referrals for child protective services and self-report^67,68 (both [100%] studies conducted by one group). After legalization, there was a significant positive association between density of medical cannabis dispensaries and referrals for child protective services, though this became nonsignificant after adjusting for density of alcohol outlets and neighborhood factors⁶⁸; the association was significant for parents’ self-reported physical abuse, but not supervisory or physical neglect⁶⁷. Based on classification of evidence for these 2 studies, there is insufficient evidence regarding effects of legalization on child abuse and neglect (2 Class IV studies^67,68). Moreover, risk of bias assessment judged these studies at critical risk of bias due to confounding (see Figure 2).

Child Outcomes

Ten studies examined unintentional pediatric cannabis exposure, assessed using reports to poison control centers and urine toxicology at hospital visits^69–78‡ (six [75%] studies conducted by two groups^72–77). There were no studies examining the effects of legalization on child outcomes other than pediatric cannabis exposure, such as child behavioral adjustment. There was evidence of a significant increase in pediatric cannabis exposure in the 6 studies using the strongest designs^{70,72–74,76,78‡} (i.e., pre-post legalization without state comparison). One study found increases in pediatric cannabis exposure continued to increase during the post-legalization period⁷⁷. Based on classification of evidence for these 10 studies (6 Class II studies^{70,72–74,76,78‡}), it is likely legalization increases pediatric cannabis exposure. However, risk of bias assessment judged even the studies using the strongest designs at serious to critical risk of bias due to confounding (see Figure 2).

Discussion

Rapid shifts in medical and recreational cannabis legalization in the United States have raised concerns of potential harm to minors and prompted recommendations against cannabis use⁷ and use during pregnancy and breastfeeding⁸. However, these recommendations draw from a relatively sparse and critically unexamined literature. We conducted a systematic review of studies examining legalization and parental cannabis and other substance use, parenting, and young children, and assessed classification of evidence and risk of bias using established rating systems^37,38. Key conclusions are that it is likely legalization increases maternal cannabis use during pregnancy and postpartum, parental use, and approval of adult use. In contrast, there is insufficient evidence legalization leads to maternal use of other substances during pregnancy and postpartum, parental use of other substances, or changes in parental attitudes toward use. It is possible legalization increases some adverse perinatal outcomes, though findings were inconsistent. It is likely legalization increases unintentional pediatric cannabis exposure, but there is insufficient evidence regarding cannabis storage in the home, parental approval of adolescent use, parenting, child abuse and neglect, or other child outcomes, including child behavioral adjustment. Pediatric cannabis exposures are associated with more emergency department visits than typical poisoning exposures reported to poison centers (93% versus 9%; Wang et al., 2013), and cannabis legalization is associated with increases in exposures leading to moderate or major medical outcomes (as defined by the National Poison Data System⁷¹) and critical care admissions^36,75§,74. Concerns have been noted regarding edible cannabis products, which come in forms that are attractive to children and can contain high amounts of THC. Edible products are responsible for a majority of pediatric exposures (48% to 69%^69,70,76,77), and were an independent predictor of ICU admission³⁶. Taken together, results of this systematic review suggest legalization does lead to increased cannabis use among parents, including maternal cannabis use during pregnancy, and parents’ approval of cannabis use among adults. However, with the notable exception of unintentional pediatric cannabis exposure, it is not clear that this increased parental cannabis use and approval of cannabis use is associated with an increase in negative parenting or maladaptive child outcomes.

Does Parental Cannabis Use in a Legal Context Have Causal Effects on Children and Families?

The primary concern regarding normalization and increasing cannabis use among parents is that parental use may have negative implications for minors. The AAP opposes legalization and “discourages the use of cannabis by adults in the presence of minors because of the important influence of role modeling by adults on child and adolescent behavior”⁷. Caution regarding legalization may be warranted if legalization leads to increased parental modeling of use (as the AAP suggests), reduced parental monitoring, and parenting dysfunction, which in turn lead to negative child outcomes. The results of the systematic review suggest more research is necessary before such conclusions can be made. Although legalization likely increases parental use, there was insufficient literature on the effects of legalization on parenting or child abuse and neglect, or child outcomes, including behavioral adjustment. Moreover, we are unaware of any studies examining whether effects of parental use on parenting and children varies across legal versus illegal contexts.

Parents who use cannabis in an illegal context, running the risk of serious negative consequences (e.g., loss of employment, children removed from the home, arrest and imprisonment), likely have greater propensity toward problematic use, tendency toward delinquency, and associated dysfunction than parents who do not use illegally. Parents who use only in a legal context may be more responsible, occasional users (as for moderate and responsible, legal adult use of alcohol). If this were the case, we would expect to find parents’ use in a legal context is not necessarily associated with parenting impairment or negative child outcomes. Studies addressing these issues are another critical area of future research.

Another important consideration is that effects of legalization may not be uniform across parents. Genetic influences on substance use are larger in more permissive environments (e.g., permissive tobacco norms⁸⁰, in college where binge drinking is common⁸¹). Thus, legalization, which increases cannabis availability and acceptability, may have the largest effects on parents with greater genetic liability toward substance use and abuse, which is also likely associated with other risk factors that might account for negative outcomes, including family dysfunction. Failing to consider implications of genetic liability and associated risk factors may result in missed opportunities to identify at-risk parents, children, and families who may benefit most from targeted prevention efforts.

Conclusions and Future Directions

This systematic review indicates evidence of some potentially causal effects of legalization on maternal use during pregnancy and breastfeeding, parental cannabis use, some adverse perinatal and postnatal outcomes, and children’s unintentional exposure to cannabis, suggesting careful consideration of effects on parents, children, and families is warranted before endorsing comprehensive legalization across the United States. However, there are still critical gaps in the existing literature. Although recent reviews document a growing literature on effects of legalization on adolescent cannabis and other substance use and adolescent attitudes toward use^19,23–28, to our knowledge, there have been no studies examining effects of legalization on aspects of parenting other than child abuse and neglect or on child behavioral adjustment. We have emphasized the strengths of the DiD approach for identifying casual effects of legalization, but other, complementary causally informative study designs are also informative. Although it is not feasible to randomly assign parents, children, or families to states with and without cannabis legalization, as in an experiment, experimental animal studies and genetically informative twin family studies have provided insights into causal effects of cannabis and other substance use for important aspects of functioning^82–85. Additional causally informative research using epidemiological samples and quasi-experimental designs is clearly needed. Three key questions are: (1) Does legalization cause parenting impairment, family dysfunction, and negative child outcomes? (2) Does legalization increase these outcomes among already using parents? and (3) Does legalization increase these outcomes among parents who use responsibly and moderately, and only in a legal context?

This systematic review was limited to studies conducted in the United States, as we identified only one study in the full-text review conducted outside the United States³⁵**. Most of the included studies did not consider race or ethnicity as a potential moderator, with one notable exception finding the increase in parental cannabis use following medical legalization was strongest for non-Hispanic white individuals, and no significant race or ethnic differences following recreational cannabis legalization³¹. Additional studies addressing this issue are needed, as effects of legal enforcement⁸⁶ and consequences of use (e.g., problematic cannabis use⁸⁷) are more severe for racial and ethnic minorities. In general, lack of generalizability is a significant limitation in the literature.

Public policy guidelines and preventive-intervention efforts that focus exclusively on cannabis use ignore broader social and individual factors that may account for negative outcomes among adolescents and adults, as well as parents, children, and families. In the absence of solid evidence of causal effects of legalization on parents, parenting, children, and family, and of parental use on children’s development and functioning, it is critical researchers and clinicians are aware that cannabis use and abuse may reflect risk, or liability toward substance use and abuse, that should be carefully assessed and addressed.

Highlights.

• Cannabis legalization leads to increased cannabis use among adults

• Legalization may have negative implications for minors via effects on parents

• Critical lack of causally informative studies with parents and young children